Rethinking Chinarsquos Coercive Capacity AnExamination of PRC Domestic SecuritySpending 1992ndash2012Sheena Chestnut Greitens

AbstractDiscussions of Chinarsquos rising domestic security expenditure often present thisspending as evidence of the Chinese Communist Partyrsquos strong coercive cap-acity This article argues that a lack of theoretical clarity about domestic secur-ity has resulted in flawed conclusions about these expenditures and theirimplications for Chinarsquos coercive capacity Challenging the conventional wis-dom the article analysesChinarsquos domestic security spending from1992 through2012 and argues that it is important to consider not only the total amount thatChina spends but also how it spends these resources and the magnitude of thethreats that this expenditure must address It finds that Chinarsquos domestic secur-ity spending is not historically unprecedented is not expanding as a proportionof national expenditure and is not necessarily high (or producing high coercivecapacity) when compared to other countries The article also shows that certainlocations struggle more to fund their coercive capacity than others and thatthese locations overlap with areas where internal security threats may beparticularly acute The challenges that the coercive apparatus must addresshave also grown over the same period during which domestic security spendinghas risen Finally attempts to improve the political position of Chinarsquos coerciveagents cannot be equated with improvements in their capacity to manageChinese society Cumulatively this reassessment provides more evidence ofthe limitations on Chinarsquos coercive capacity than of its strength

Keywords domestic security internal security public security policingauthoritarian stability coercive capacity stability management (weiwen)Commission on Law and Politics (zhengfawei) China

InMarch 2011 international news outlets reported that internal security spending inChina had for the first time surpassed external defence expenditure1 Double-digitincreases in spendinghadpushed thedomestic securitybudgetupwardat anexponen-tial rate as shown in Figure 1 and the internal security budget remained higher thanthe defence budget for several subsequent years (see Table 1)

Department of political science University of Missouri Email greitenssmissouriedu1 Buckley 2011 2012a Blanchard and Ruwitch 2013

1

copy SOAS University of London 2017 doi101017S0305741017001023

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S0305741017001023Downloaded from httpswwwcambridgeorgcore University of Missouri-Columbia on 05 Jul 2017 at 203959 subject to the Cambridge Core

Since then domestic security has occupied a prominent place in publicand academic discussions of Chinese politics and society and the domesticsecurity budget is a frequent point of reference2 But how should observersinterpret this apparently astonishing growth Journalists commonly suggestthat Chinarsquos leaders are raising spending in an unprecedented fashion toaddress heightened insecurity and as part of an increasing emphasis on ldquostabilitymaintenancerdquo (weihu wending 维护稳定 or weiwen 维稳) in officialdiscourse and behaviour3 Academic literature however has taken a differenttack citing spending increases as evidence of the ldquostrong coercive capacityrdquo ofthe Chinese state following a ldquodramatic expansionrdquo in that capacity since theearly 1990s4

This article demonstrates that neither of these perspectives is entirely correct inlarge part because both rest on an overly simplistic idea of what ldquocoercive cap-acityrdquo means and because they selectively employ indicators that are not agood match for the concept In the pages that follow I seek to clarify the discus-sion on Chinarsquos internal security spending in order to shed light on the role ofcoercive capacity in Chinarsquos authoritarian stability This is done in two waysfirst by carefully defining coercive capacity and then by using new theoretically

Figure 1 Chinarsquos Total Domestic Security Expenditure 1992ndash2012

SourcesMOF 1992ndash2002 NBS 2003ndash2013

2 ldquoEverything Xi wantsrdquo The Economist 4 July 2015 Lampton 2015 Erickson and Liff 20163 On the history and development of weiwen see Yuen 2014 Kan 2013 Qian 20124 Wang 2014a 2014b Wang and Minzner 2015 See also Chen 2013 Xie and Shan 2013 SDRG 2010

2 The China Quarterly pp 1ndash24

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S0305741017001023Downloaded from httpswwwcambridgeorgcore University of Missouri-Columbia on 05 Jul 2017 at 203959 subject to the Cambridge Core

Table 1 Chinarsquos External Defence and Internal Security Spending 2010ndash2013

Year 2010 2011 2012 2013Defence budget 532115 billion yuan

(US$846 billion)601156 billion yuan

(US$9558 billion)670274 billion yuan

(US$10657 billion)740622 billion yuan

(US$11775 billion) increase 1297 115 105Defence expenditure 533337 billion yuan

(US$8479 billion)602791 billion yuan

(US$9584 billion)669192 billion yuan

(US$10639 billion)741062 billion yuan

(US$11782 billion) increase 1302 1102 1074Int security budget 514007 billion yuan

(US$8172 billion)624421 billion yuan

(US$9928 billion)701763 billion yuan

(US$11157 billion)76908 billion yuan

(US$12228 billion) increase 2148 1239 959Int security expenditure 55177 billion yuan

(US$8773 billion)630427 billion yuan

(US$10023 billion)71116 billion yuan

(US$11306 billion)778678 billion yuan

(US$12379 billion) increase 1426 1281 949

SourcesldquoQuanguo gonggong caizheng zhichu juesuan biaordquo (National public expenditure final account tables) published by theMinistry of Finance at httpyssmofgovcn2010juesuan201107t20110720_578444html httpyssmof

govcn2011qgczjs201207t20120710_665233html httpyssmofgovcn2012qhczjs201307t20130715_966261html httpyssmofgovcn2013qgczjs201407t20140711_1111874htmlNotes

In 2014 and 2015 the Chinese government declined to release the total amount spent on internal security All US$ estimates adjusted based on 2013 annual average exchange rates

RethinkingChinarsquos

CoerciveCapacity

3

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S0305741017001023D

ownloaded from

httpsww

wcam

bridgeorgcore University of M

issouri-Columbia on 05 Jul 2017 at 203959 subject to the Cam

bridge Core

derived indicators placed in appropriate comparative perspective to provide are-assessment of Chinarsquos coercive capacity5

The revised analysis of Chinarsquos domestic security budget demonstrates thatconventional wisdom exaggerates the exceptional and unprecedented nature ofChinarsquos increases in spending on domestic security The Chinese CommunistParty (CCP) has indeed attempted to strengthen its coercive capacity but it isprobably doing so because it perceives that capacity to be inadequate for man-aging Chinarsquos rapidly changing society In other words budget trends in thelast two decades likely indicate the weakness and limitation of CCP coercive cap-acity during this period not its strengthThis article proceeds in four sections The second section establishes the importance

of coercive capacity outlines the current lack of clarity in its theorization and meas-urement andoffers a theoretically guided redefinitionof the concept and some sugges-tions for better measurement The third section offers a revised analysis of Chinarsquosdomestic security spending based on the reconceptualization offered in section twoThe fourth section concludes by discussing the implications of this analysis for currentdevelopments It discusses how the framework proposed heremay usefully illuminateChinarsquos internal security behaviour in recent years including the reorganization ofdomestic security forces the creation of newnational security legislation and the tigh-tening of control over Chinese society under Xi Jinpingrsquos习近平 leadership

Assessing Chinarsquos Coercive CapacityCoercive capacity has long been recognized as a critical component of authoritar-ian stability6 Few studies however examine how that capacity is generated andsustained or assess the relative importance of the budget in that process This sec-tion asks what is coercive capacity and how should it be measured More specif-ically it discusses how internal security expenditures are related (or not) tocoercive capacity both generally and in the China caseThe China field is divided on these questions The dominant interpretation sug-

gests that since 1989 China has been undergoing a long-term process of ldquosecuri-tizationrdquo (of which weiwen is the most recent manifestation) that has strengthenedthe regimersquos coercive capacity7 This explanation points to increased spending onthe coercive apparatus as a major piece of evidence for its claims8 Even worksthat are normatively critical of the increased spending interpret it as an indicationof rising coercive capacity9

5 On comparative perspective see Reny 2011 Johnston 2012 OrsquoBrien 2011 On coercive capacity seeGuo Xuezhi 2012 Scoggins and OrsquoBrien 2016 Deng Yanhua and OrsquoBrien 2013 Tanner andGreen 2007 On Chinarsquos authoritarian resilience see (among many others) Nathan 2003 Pei 2012

6 Bellin 2005 Brownlee Masoud and Reynolds 2015 Greitens 2016 Pei 2012 32 Skocpol 1979 Wang2014a

7 The term ldquosecuritizationrdquo is used in Wang and Minzner 20158 The other is the promotion of police officials within the Party hierarchy addressed below Wang and

Minzner 2015 Wang 2014a 2014b9 Chen 2013 Xie and Shan 2013 SDRG 2010

4 The China Quarterly pp 1ndash24

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S0305741017001023Downloaded from httpswwwcambridgeorgcore University of Missouri-Columbia on 05 Jul 2017 at 203959 subject to the Cambridge Core

A few Chinese-language studies however adopt a more sceptical tone notingthat the 1994 fiscal reforms exacerbated local budget problems even in the con-text of increased overall spending They argue that the Ministry of PublicSecurityrsquos (MPS) frontline officers often have inadequate resources for thetasks they are expected to perform10 Articles in Chinarsquos public security journalscommonly discuss how to deal with the negative consequences of budgetaryshortfalls11 and how to maximize efficiency given limited resources12 Thesecomplaints are not necessarily to be taken at face value ndash under-resourcing isafter all a perennial complaint of bureaucrats the world over ndash but neither shouldthey be dismissed out of hand Instead this article looks at the disjuncturebetween these two perspectives and asks what is Chinarsquos coercive capacity Toanswer that question it is necessary to generate theoretically appropriate mea-sures of coercive capacity and use these to judge where China fallsAnalysing spending is attractive for many reasons not least because it is quan-

tifiable13 Rigorous assessments of domestic coercive capacity and of its financialunderpinnings however have been hampered by three key challenges each ofwhich directly affects the debate over China First is a simple lack of transparencyand data There is no comprehensive dataset for internal security spending andassets comparable to the military compendia published annually by theInternational Institute for Strategic Studies (IISS) and the StockholmInternational Peace Research Institute (SIPRI) Both military and internal secur-ity budgets are sensitive subjects but the comparative lack of external pressure tomake domestic expenditures transparent combined with the institutional hetero-geneity of the internal security apparatus compared to the military hinders rigor-ous interpretation14

Second differentiating internal from external security is often difficult espe-cially when assets or personnel are fungible or dual-use As a result there is noconsensus on how to make this demarcation datasets on military expenditureoften include organizations with a domestic focus but at the same time excludeactors that have a large international impact without providing a justification15

The IISS and SIPRI for example include the Peoplersquos Armed Police (PAP) ndash an

10 Xie 2013a 2013b Scoggins and OrsquoBrien 2016 See also Guo Gang 2012 Luuml and Landry 2014 Whiting2004

11 Yao 2004 Ye 2006 Deng Xuan 201112 Xie and Dang 2013 Shi Xiaochen and Zhang 201513 Andreas and Greenhill 201014 Militaries are generally cross-nationally comparable in terms of having recognizable service branches In

domestic security however each country tends to create its own mix of nationallocal police intelligenceagencies presidentialstate security agencies courts etc For a comparative approach to domestic secur-ity bureaucracies see Greitens 2016 Expenses may also be funded off-budget both generally and in theChinese case Analysts disagree on how large Chinarsquos extrabudgetary expenses on domestic securityespecially funds earmarked for weiwen (weiwen jingfei) are likely to be I acknowledge that unobservedextrabudgetary spending may introduce downward bias on the data but believe the data is still valuableso long as appropriate caveats are provided

15 On how this lack of consensus affects military spending estimates see Liff and Erickson 2013 Forsythe2014

Rethinking Chinarsquos Coercive Capacity 5

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S0305741017001023Downloaded from httpswwwcambridgeorgcore University of Missouri-Columbia on 05 Jul 2017 at 203959 subject to the Cambridge Core

organization strengthened post-1989 to take over domestic security from the PLA ndash

in Chinarsquos defence spending but they exclude maritime law enforcement agenciesthat operate in disputed territorial waters (such as the South China Sea)16 Onthe other hand studies of authoritarian politics typically use military spendingas a proxy for coercive capacity17 in Chinarsquos case this excludes the mainorganizationsspending tasked with responsibility for domestic security whichis nonsensical if internal security is the concept of theoretical interestSeparating law enforcement and criminal justice from political policing is the

third challenge The extent to which normal judicial-legal institutions are usedfor political policing and how exactly they are employed varies widely acrosscountries and across time18 Discussion of Chinarsquos domestic security budgetoften treats this spending as aimed entirely at suppressing political oppositionto the CCPrsquos single-party rule commonly citing the growth of ldquomass incidentsrdquoto explain budget increases Yet in fact in China a single budget and organiza-tional system ndashthe political-legal system (zhengfa xitong政法系统) ndash address bothcriminal and political aspects of domestic security At the local level censorsremove both pornography and political commentary and MPS offices handlecrime control as well as protest management19 The ldquointernal security budgetrdquosupports law enforcement and criminal justice functions that would still requirefunding even if China democratized tomorrow Yet discussions of Chinarsquosdomestic security budget seldom consider whether crime rather than politicalopposition has played any role in the recent budget increasesThe above paragraphs highlight the risk of uncritically employing budget sta-

tistics to gauge Chinarsquos domestic coercive capacity What then should analystsuse instead I suggest not that budget statistics should be abandoned but that theymust be interpreted more carefully in historical and cross-national comparativecontext to make judgments about their importance for ldquocoercive capacityrdquo20

Specifically drawing on recent findings in security studies I argue that anyassessment of coercive capacity must go beyond simply what a country spendsto incorporate two additional factors what that money is spent on and what itis spent againstIt is important to consider what domestic security budgets are spent on because

two countries with equivalent budgets may choose to spend that money in waysthat make their expenditures more or less effective Studies of military

16 PAP spending is included in Chinarsquos statistical yearbooks as a major category under domestic securityPrior to March 2013 maritime law enforcement was handled by five agencies all civilian post-consolidation responsibility lies with the State Oceanic Administration (under the Ministry of Landand Natural Resources) The MPS also issued passports in 2012 reportedly without consulting theForeign Ministry that showed disputed islands as Chinese territory Blasko and Corbett 1998Cheung 1996 Erickson and Collins 2013 Liff and Erickson 2013 Forsythe 2014 Fravel 2007Goldstein 2010 Jakobson 2014 Martinson 2014 Ruwitch 2012 Tanner 2002 IISS 2001ndash2012Wines 2009

17 Bellin 2005 31 Albertus and Menaldo 201218 Solomon 2007 Greitens 2016 Scoggins and OrsquoBrien 201619 King Pan and Roberts 2014 Scoggins and OrsquoBrien 201620 On the value of comparison see Liff and Erickson 2013 Johnston 2013 34

6 The China Quarterly pp 1ndash24

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S0305741017001023Downloaded from httpswwwcambridgeorgcore University of Missouri-Columbia on 05 Jul 2017 at 203959 subject to the Cambridge Core

effectiveness (where the acquisition of hardware arguably provides better evi-dence of capacity than it does for internal security) have shown that the correl-ation between spending and performance is tenuous they conclude that ldquoitrsquosnot what states spend itrsquos what they do with what they spendrdquo that matters21

Just as a military that invests primarily in tanks will find itself disadvantagedin fighting a predominantly naval war domestic security forces that are trainedand equipped for rural counter-insurgency may perform poorly in urban riot con-trol These studies further suggest that performance is not simply a matter of buy-ing the right equipment or training organizational attributes such asfragmentation social cohesion information management and promotion pat-terns all affect a countryrsquos ability to translate spending into military power22

Recent literature suggests that the same is true of domestic security where auto-crats face organizational trade-offs between optimizing their forces to address dif-ferent types of domestic security challenges each of which they must navigatesuccessfully in order to stay in power23 To understand whether increased spend-ing is actually increasing Chinarsquos coercive capacity then it is important to con-sider whether the organizations that receive that spending are effectivelyemploying it for the purposes of controlling Chinese societyA useful definition of coercive capacity and its importance for authoritarian

rule also requires an understanding of what the budget is being spent againstIn other words how does the coercive apparatusrsquo ability measure up againstthe challenges it is expected to handle24 It makes little intuitive sense to claimthat the coercive capacity of (for example) a 500-person police force with a $1million budget would be the same in a city of 20000 as it would be in a cityof 2000000 or that it would have the same capacity to keep order in a citywith extremely high crime and violence as in a city where crime rates are muchlower During the period analysed here Chinese society changed tremendouslythe population grew both crime and incidents of political protest increased in fre-quency and many of the traditional institutions of social control that character-ized Maoist China were weakened or abolished The CCPrsquos capacity to enforce itsrule and stay in power depends not just on its raw spending or even on its abso-lute ability but on its ability relative to the also changing ability of Chinese soci-ety to challenge it Coercive capacity will only be a useful predictor of regimesurvival if it is relative in its conceptualization and measurementTheoretically it only makes sense to equate spending with effective coercive

capacity (especially if coercive capacity is then to be credited with regime sur-vival) if how that spending is employed and the magnitude of the challenges itmust address are also considered The sections that follow show that careful ana-lysis of Chinarsquos domestic security budget placed in historical and cross-national

21 Biddle 2006 Brooks and Stanley 2007 Grauer and Horowitz 2012 Talmadge 201522 Horowitz 2010 Narang and Talmadge 201723 Greitens 2016 Roessler 201124 Here my argument parallels a long-standing claim in international security that relative rather than

absolute gains are what matter for inter-state conflict Grieco 1993

Rethinking Chinarsquos Coercive Capacity 7

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S0305741017001023Downloaded from httpswwwcambridgeorgcore University of Missouri-Columbia on 05 Jul 2017 at 203959 subject to the Cambridge Core

comparative context and assessed alongside the above criteria undercuts theclaim that increased spending has created an increase in CCP coercive capacityRather the data more likely indicate weakness or limitation ndash a finding that mayalso more logically explain Chinarsquos recent domestic behaviour

Reframing Chinarsquos Internal Security SpendingThe following section outlines a revised interpretation of Chinarsquos domestic secur-ity spending offering several correctives to academic and conventional wisdomFirst it looks at how much China is spending in historical perspective showingthat although total spending has increased domestic security has remainedroughly constant as a proportion of national expenditure over time Second itexamines what China spends its domestic security budget on ndash what categoriesand what regions ndash to show that Chinarsquos spending and the coercive capacity itbuys is not necessarily exceptional in cross-national terms and may even be fairlylow Third it investigates what China spends its budget against showing that thecombination of rising crime and increasing levels of political protest suggests thatthe challenges facing the coercive apparatus may well be outstripping its sup-posed increases in capacity Finally it presents an organizational analysis show-ing that efforts to raise the political power of the coercive apparatus are not thesame as strengthening its ability to manage society Cumulatively these pointssuggest that increased overall spending on domestic security is likely to indicatendash and be motivated by ndash the inadequacy of Chinarsquos coercive capacity rather thanits repressive strength

Figure 2 Internal Security Expenditure as Proportion of Total Expenditure

SourcesMOF 1992ndash2002 NBS 2003ndash2013 see also Guo Gang 2012 Pre-1997 statistics omit prisons

8 The China Quarterly pp 1ndash24

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S0305741017001023Downloaded from httpswwwcambridgeorgcore University of Missouri-Columbia on 05 Jul 2017 at 203959 subject to the Cambridge Core

What China spends on domestic security (historical perspective)

Chinarsquos domestic security budget is most often described using percentageincreases from the year before or in comparison to the countryrsquos defence budgetBoth of these metrics give the impression that domestic security spending hasrecently increased on an unprecedented and dramatic (ldquodouble-digitrdquo) scaleand that this spending is consuming an increasingly large chunk of the resourcesof the Peoplersquos Republic of China (PRC) Neither claim is accurate Chinarsquosentire budget has been rising fast producing double-digit growth in most categor-ies The growth in aggregate health care expenditure is as exponential as domesticsecurity spending growth in social security spending has like domestic securityoutstripped growth in defence spending since the early 1990s25 More than thatsince the 1980s the PRC has shifted an increasing share of its budget towardseducation health care social security and housing26 Domestic security spendingrsquosshare of total expenditure however has stayed relatively constant between 5 and7 per cent of total expenditure (Figure 2)The figures presented in this article draw on Chinarsquos official statistical year-

books Adjustments have been made where necessary in order to ensure thatthe aggregate figures include comparable sub-categories over time (For examplelabour re-education was sometimes included in the aggregate yearbook figuresometimes listed separately here it is included in totals regardless of where inthe yearbook it appeared) The percentage of national expenditure allocated todomestic security ranged from a low of 44 per cent in 1992 to a peak of 70per cent in 2007 declining to 56 per cent in 2013China is spending more money on everything not just on domestic security

domestic security is not getting a bigger share of the pie now than before Thissuggests that to understand increased domestic security spending it is best tostart with what is driving overall budget increases ndash often attributed to factorslike increasing personnel costs ndash rather than assuming that domestic security issomehow exceptional27 Indeed the pattern here suggests that the causal forcesresponsible for spending increases are not in fact either unique to internal secur-ity nor particularly new since the percentage spent on domestic security has notdramatically increased in recent years if anything it has declinedOne potentially complicating factor is that under the ldquosecuritizationrdquo of the

Chinese state more parts of the political system (including bureaucrats responsiblefor everything from labour to the environment) now share responsibility for ldquostabil-ity maintenancerdquo but do not appear in the domestic security budget Typicallyhowever the responsibility of these actors in terms of stability maintenance is

25 This holds even using high-end estimates of military spending from SIPRIIISS Sheen 2013 StateCouncil Information Office 2012

26 Zhu and Wang 201127 Unfortunately the data necessary to fully test this hypothesis do not (yet) exist Qualitative research sug-

gests regional disparities in police salaries are consistent with the subnational variation analysed hereFor example police in Guangdong earn 6ndash7 times more than police officers in many other provincesScoggins 2016

Rethinking Chinarsquos Coercive Capacity 9

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S0305741017001023Downloaded from httpswwwcambridgeorgcore University of Missouri-Columbia on 05 Jul 2017 at 203959 subject to the Cambridge Core

preventive to minimize societal and citizen grievance and forestall unrest I focushere on a somewhat narrower definition of coercive capacity and restrict the ana-lysis to the set of actors who exercise and implement the regimersquos monopoly on(physical) force rather than include all those who are responsible for the broaderpolitical imperative of reducing citizensrsquo grievances with the state or regime

How China spends its domestic security budget (categories and geography)

Coverage of Chinarsquos internal security spending figures seldom discloses preciselywhat these statistics include or how they compare to other countries Figure 3shows the categorical allocation of Chinarsquos domestic security spending Thebulk of Chinarsquos domestic security budget since 1996 has gone to the Ministryof Public Security (gongrsquoan 公安) ranging between 588 per cent (2009) and 632per cent (1996) Other major categories each year include the PAP (wujing 武警)national security (guojian anquan 国家安全) procuratorate ( jiancha 检察) courts( fayuan 法院) Ministry of Justice (sifa 司法) prisons ( jianyu 监狱) andre-education through labour (laojiao劳教) Since 2006 the budget has also includedcategories for protection of state secrets (guojia baomi 国家保密) anti-smuggling

Figure 3 Categories of Domestic Security Spending by per Cent of Budget1996ndash2009 Excluding MPS

SourcesMOF 1996ndash2009

NotesFor 1996ndash1997 the yearbooks did not include prison and labour re-education figures in total domestic security spending (but

included them in subsequent years) To make the data comparable these sub-categories were added to the 1996ndash1997 totals sothe estimate of total internal security spending is higher than that in the yearbooks

10 The China Quarterly pp 1ndash24

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S0305741017001023Downloaded from httpswwwcambridgeorgcore University of Missouri-Columbia on 05 Jul 2017 at 203959 subject to the Cambridge Core

police ( jisijing 缉私警) and ldquootherrdquo (qita 其他) although each of these is fairlysmall Most categories show a fair amount of stability over time the budget per-centage going to prisons and labour re-education declined the most28 while thebiggest spending increases were on courts and the PAPAs noted above this budget includes not only explicitly ldquopoliticalrdquo organiza-

tions such as state security and the PAP but also institutions with broader crim-inal justice functions such as local police and courts Is Chinarsquos spending on thissystem exceptional To construct a preliminary answer to this question I aggre-gated budgets for comparable institutions in the United States and Russia ndash twocountries that like China are great powers with a large territory diverse geog-raphy and significant internal security concerns either criminal or political(but which as a robust and a weak democracy respectively might plausiblyspend less on internal security than Chinarsquos fully authoritarian system)29 Onlyknown and measurable costs are included making the estimates conservativeTable 2 shows this comparison for 2013When roughly equivalent categories are compared China spent less than the

United States on domestic security for a larger territory and much larger popu-lation and that both China and Russia spent roughly comparable amounts ondomestic and external security The United States spent around $489 per capitaon domestic security while Russia spent $393 China spent approximately $9230

Until a full cross-national dataset on domestic security spending is availablethere is no way to tell how these three countries stack up against ldquothe averagerdquoin their budgetary allocations for internal security or even against various com-parison categories that might be of theoretical interest such as great powersauthoritarian regimes communist countries etc These illustrative data howevershould call into question the assumption that China is an exceptionally heavyspender on domestic security as often implied or that Chinarsquos high spendingis simply the consequence of its authoritarian system

Table 2 Comparison of US Russia and PRC Security Spending 2013 (US$)

Country Defencespending

Domestic securityspending

Domestic securityspending (per capita)

US $5266 billion $155 billion $489Russia $634 billion $56 billion $393China $120 billion $124 billion $92

SourcesDOD 2013 Cooper 2014 Janersquos lists Russiarsquos 2013 defence spending as $688 billion for 2013 Calculations based on US popu-

lation of 317000000 Russian population of 142355000 PRC population of 1355000000 US Census Bureau wwwcensusgovpopclock and httppressihscompress-releaseaerospace-defense-terrorismglobal-defence-budgets-overall-rise-first-time-five-yearsFor an explanation of construction of the US spending estimate see Appendix 1

28 Note that these data end prior to the official abolishment of labour re-education in December 201329 Polity scores range from minus10 (full autocracy) to 10 (full democracy) The US Polity IV score in 2013 was

10 Russiarsquos score was 4 Chinarsquos score was minus830 On the use and misuse of per capita figures see Xiao 2013

Rethinking Chinarsquos Coercive Capacity 11

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S0305741017001023Downloaded from httpswwwcambridgeorgcore University of Missouri-Columbia on 05 Jul 2017 at 203959 subject to the Cambridge Core

Perhaps more importantly Chinarsquos lower spending also results in a smaller coer-cive presence deployed on the ground PRC domestic security spending is not lowersimply because coercive capacity ndash for example the cost of hiring a police officer ndash ischeaper inChina andBeijing is not buyingmore coercive capacity for a lower priceIt has fewer police per capita than theUS at 138 officers per 1000 residents in 2009(the last year for which an estimatewas available) compared to aUS average of 23and a Russian average of around 531 In fact China has a lower per capita policeratio than many other countries (see Figure 4)These data suggest that complaints about Chinarsquos police shortfall in public

security journals and Chinese media and the recruitment of volunteers to fillthose shortfalls are not simply the result of bureaucratic dissatisfaction and pos-turing for publicity (although this may also be the case)32 China is not gettingmore for its money it is actually getting less coercive power as a result oflower spendingAnalysing the geographic distribution of Chinarsquos domestic security spending

similarly suggests that the decentralization of domestic security budgets mayhave weakened Chinarsquos coercive capacity particularly in areas perceived to beresistant to CCP rule Previous analyses have noted the dominance of provincialand local spending relative to that of the central government Figure 5 shows thatthis trend has deepened over time Indeed the percentage of internal securityexpenditure funded by local rather than central coffers rose significantly from1992 (687 per cent) to 2012 (834 per cent) This trend contrasts with the defencebudget where around 85 per cent of spending is central and the shift towardslocal expenditure continued even after the 2003 reforms which were aimed atstrengthening central control by increasing transfer payments (zhuyi zhifu 转移

支付) to local public security departments33 The transferred funds are intendedto prevent local departments from levying excessive and unpopular fines to coverbudgetary shortfalls but their usage is restricted to certain categories whichoften leaves local government with a heavy burden In China where central over-sight is often framed as the answer to local abuse and predation the gradualweakening of central financial control over coercion is notable and consistentwith the idea of China as a ldquofragmented authoritarianrdquo polity34

Past studies have shown the importance of local financial capacity for deter-mining localitiesrsquo domestic security spending wealthier eastern provinces spend

31 Reaves 2010 2011 Scoggins and OrsquoBrien 2016 Scoggins 201632 Rising salarypersonnel costs would explain both increasing expenditure and a shortfall in police person-

nel Zhong 2009 Hu 2009 Bureau of Justice 2013 ldquoZhongyang zongzhiban quntixing shijian ji xinfangzongliang xiajiang mubiao shixianrdquo (Central CMPS Commission reduction targets of mass incidentsand petitions realized) China Net 6 February 2009 ldquoJiceng minjiang mianlin zuida de kunnan shi jinglibuzurdquo (The biggest problem with civilian police is the shortage of police) Renmin Net 9 March 2013ldquoChinarsquos police complain of manpower shortage in countryside despite crime rate fallingrdquo Xinhua 15November 2006

33 Xie 2013a 82ndash85 90 Tanner and Green 200734 Tanner and Green 2007 Mertha 2009 Lu and Landry 2014 Wallace 2014 Lampton 1987a 1987b

Lieberthal and Oksenberg 1988 Lieberthal and Lampton 1992

12 The China Quarterly pp 1ndash24

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S0305741017001023Downloaded from httpswwwcambridgeorgcore University of Missouri-Columbia on 05 Jul 2017 at 203959 subject to the Cambridge Core

Figure 4 Global Police per Capita Ratios

SourceUN Office on Drugs and Crime 2009 ldquoTotal police personnelrdquo httpsdataunodcorglf=1amplng=en Multiple years were tested

and comparable results found

Rethinking Chinarsquos Coercive Capacity 13

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S0305741017001023Downloaded from httpswwwcambridgeorgcore University of Missouri-Columbia on 05 Jul 2017 at 203959 subject to the Cambridge Core

more than poorer inland ones in both gross and per capita terms35 Guangdongfor example has the largest domestic security budget of any province whileNingxia consistently has the lowest Guangdongrsquos spending per capita is almostthree times that of Ningxia Poorer inland provinces and regions however dospend a higher proportion of their revenue on domestic security and(post-2003) fund more of their budgets through central transfers36 In otherwords provinces that can spend more on domestic security do so those that can-not receive money from the centre to help offset perceived shortfalls This sub-national variation in the financial foundation of coercive capacity hidden bythe more common references to annual percentage increases and defence budgetcomparisons is consistent with a strain on the coercive apparatus rather thanevidence of robust capacityImportant for assessing the regimersquos coercive capacity relative to society that

strain ismore pronounced in particular areas Figure 6 shows that per capita spend-ing on domestic security has increased more steeply in some regions than in othersespecially in the latter half of the 2000s The two most noticeable increases are inBeijing ndash unsurprising given the presence of the top leadership ndash and Tibet where

Figure 5 Local Spending as Proportion of Total Internal Security Spending

SourceMOF 1992ndash2002 NBS 2003ndash2013

35 Xie 2013a 8636 Ibid Elsewhere however Xie says that per capita spending in more developed wealthier provinces is

lower eg Qinghai spent 368 yuan per person in 2008 on domestic security while Shandong spent184 yuan Xie 2012 24

14 The China Quarterly pp 1ndash24

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S0305741017001023Downloaded from httpswwwcambridgeorgcore University of Missouri-Columbia on 05 Jul 2017 at 203959 subject to the Cambridge Core

per capita domestic security spending started low but increased dramaticallyaround 2006 Indeed Figure 7 shows that Tibet is particularly ill-equippedfinancially to deal with challenges to CCP rule Domestic security spending relativeto GDP has always been higher in Tibet than in other provinces but the ratioskyrocketed after 2006 Although scholars have previously noted the uniquenessof central subsidies to Tibet even when compared to other poor areas in westernChina these figures add an additional layer of meaning37 In the eyes of Chinarsquosleadership Tibet represents a unique intersection high domestic security threatcombined with low financial capacity to address that threat

The threats facing China rising crime and political protest

The example of Tibet illustrates why it is useful to consider the capacity ofChinarsquos coercive apparatus relative to the challenges that the apparatus mustaddress Systematic consideration of these challenges combined with the budgettrends outlined above highlights the weakness of Chinarsquos coercive capacity ratherthan its strengthThe internal security budget covers crime control and management of political

unrest Both needs have increased steadily during the period of rising domestic

Figure 6 Domestic Security Spending per Capita over Time by Province

SourceData on spending from MOF 1996ndash2009 GDP and population data from China Data Online

37 Fischer 2015

Rethinking Chinarsquos Coercive Capacity 15

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S0305741017001023Downloaded from httpswwwcambridgeorgcore University of Missouri-Columbia on 05 Jul 2017 at 203959 subject to the Cambridge Core

security expenditure38 According to the PRCrsquos own statistics the number ofldquomass incidentsrdquo has grown steadily from 8700 in 1993 to 127000 in 2008Lawsuits against government offices and officials similarly grew from 11418in 1988 to 142861 in 2008 Both scholarly and media analyses commonly citethe rise in ldquomass incidentsrdquo to explain domestic security budget increases39

Less frequently mentioned is the fact that during the same period crime alsoincreased more than doubling from 2000 (45 million) to 2008 (9 million)Violent crimes increased at an even steeper rate The total number of criminalcases heard in court rose from 2 million in 1987 to nearly 7 million in 2008and Chinarsquos public security bureaus dealt with a far larger number of casesthan those that actually appeared in court At the same time many of the institu-tions of social control and management that existed under Mao ndash the householdregistration (hukou 户口) and work unit (danwei 单位) systems for example ndash

weakened during the process of economic liberalization and the rural-to-urbanmigration that followedGiven these changes in Chinese society and in the institutions that previously

linked the party-state with society it is not surprising that the PRCrsquos formal

Figure 7 Domestic Security Spending over Time by Province Relative to GDP

SourceData on spending from MOF 1996ndash2009 GDP and population data from China Data Online

38 Unless otherwise cited data in this paragraph are drawn from Zhongguo faluuml nianjian 1989ndash2009 ori-ginally cited in Xie 2012 5ndash7

39 Shirk 2007 57

16 The China Quarterly pp 1ndash24

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S0305741017001023Downloaded from httpswwwcambridgeorgcore University of Missouri-Columbia on 05 Jul 2017 at 203959 subject to the Cambridge Core

institutions of coercion have had to boost their resources to keep upAlthough thesedata are not conclusive evidence that the regimersquos efforts at compensation havefailed they do provide reason to question whether the supposed expansion of coer-cive capacity in China has in fact kept pace with the challenges the regime faces

Strengthening the coercive apparatus Political power versus coercive capacity

One explanation proffered for the increased resources devoted to internal security(and the decision to stop publishing the budget in 2014) was the rise and subsequentfall of Zhou Yongkang 周永康 who was minister of public security in 2002ndash2007and a member of the Politburo Standing Committee and head of the CentralPolitical-Legal Commission (zhengfawei 政法委) in 2007ndash201240 Yuhua Wangand Carl Minzner identify the incorporation and elevation of public security offi-cials within the Chinese political system under Zhou as the other main indicator (inaddition to budget increases) of the strengthening of coercive capacity since theearly 2000s political-legal committee chairs have almost always sat on provincialCCP standing committees and police chiefs are increasingly represented on provin-cial Party leadership teams41 Indeed after Zhoursquos 2002 promotion the number ofpolice chiefs involved in provincial Party leadership increased sharply from sim60per cent in 2003 to sim90 per cent by 201242 Wang also notes that police fundingis positively correlated with the rank of police chiefs and negatively correlatedwith the percentage of the labour force employed by state-owned enterprises(SOEs) He concludes that the ldquostrong coercive capacityrdquo of the CCP plays an over-looked role in Chinarsquos authoritarian resilience43

I believe these conclusions to be overstated Drawing on the previously dis-cussed literature on the organizational dimensions of coercive capacity I suggestthat these analyses have conflated two separate factors the political power ofChinarsquos coercive agents within the regime and those agentsrsquo capacity relativeto Chinese society Studies in comparative authoritarianism have previouslydocumented that intra-elite politics and societal unrest are distinct challengesto authoritarian rule and that the tools used to deal with them are differentautocrats commonly attempt to buy the loyalty of coercive agents (eitherpolitically or economically) but these efforts are often orthogonal or evencounterproductive to endowing them with the capacity to manage popularunrest44 The treatment of the coercive apparatus from 1992 to 2012 is evidencethat the CCP elevated the internal political power of coercive agents during thistime but this cannot be equated with an improvement in coercive agentsrsquo cap-acity to manage Chinese society

40 Fewsmith 2016 Buckley 2012b He 2012 Shi Jiangtao 2012 Jiang 2015 ldquoTiger in the netrdquo TheEconomist 13 December 2014

41 Wang and Minzner 201542 Wang 2014a 1743 Wang 2014a 2014b Wang and Minzner 201544 Greitens 2016 Svolik 2012 Talmadge 2015

Rethinking Chinarsquos Coercive Capacity 17

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S0305741017001023Downloaded from httpswwwcambridgeorgcore University of Missouri-Columbia on 05 Jul 2017 at 203959 subject to the Cambridge Core

In fact the trend of increasing funding for domestic security pre-dated thepractice of promoting police chiefs into the Party architecture and the improvedpolitical position of police chiefs in provincial leadership did not produce anincrease in spending on domestic security as a percentage of the overall budgetafter either 2002 or 2007 Thus even if higher-ranked police chiefs were ableto secure more funding for their provinces than lower-ranked counterparts thecoercive apparatus as a whole did not receive more funding recall fromFigure 2 that domestic security spending as a percentage of total expenditureactually declined from 2007 to 2012 In fact what the relationship betweenParty rank SOE employment and police funding suggests is that when facedwith overall budget constraints more politically powerful police chiefs shiftedresources to areas where the party-state was losing control over the labourforce ndash a pattern that as with the geographic data analysed above suggests con-cern with the coercive apparatusrsquo inadequate capacity to police Chinese societynot confidence in its strength

ConclusionDiscussions of Chinarsquos domestic security expenditure often present this spendingimplicitly or explicitly as evidence of the CCPrsquos strong and increasing coercivecapacity This article challenges that characterization by analysing not just theamount that China spends but also how it spends those resources and the mag-nitude of the threats that those resources must combat It finds that Chinarsquosdomestic security spending is not historically unprecedented not growing as aproportion of national expenditure and not necessarily producing high coercivecapacity compared to other countries It also shows that certain locations strugglemore to fund their coercive capacity than others and that these locations overlapwith areas in which internal security threats may be perceived as particularlyacute These findings are notable given that the challenges that the coerciveapparatus faces ndash in terms of both crime and political opposition ndash have grownover the same period during which spending has risen The article further findsthat it is theoretically incorrect to assume that policies that raise the politicalpower of coercive agents within the party-state are also measures that strengthentheir capacity relative to Chinese society the two phenomena are theoreticallyand empirically distinct and there is evidence for the former but not the latterCumulatively this reassessment provides stronger evidence of the limitationson Chinarsquos coercive capacity as of 2012 than of its strengthThis perspective helps to explain someof the keydevelopments inChinarsquos domes-

tic security policy since 2012 especially the creation of a largely domesticallyfocused National Security Commission the passage of new national security legis-lation and an overall tightening of social control under President Xi Jinping Thesedevelopments make the most sense if the CCP in 2012 is understood as a regimedeeply concerned about the inadequacy of its capacity to control and manageChinese society rather than a regime confident in the strength of its ability to do

18 The China Quarterly pp 1ndash24

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S0305741017001023Downloaded from httpswwwcambridgeorgcore University of Missouri-Columbia on 05 Jul 2017 at 203959 subject to the Cambridge Core

so The PRCrsquos decision not to continue releasing full budget statistics on internalsecurity after 2013 confirms the need for observers of Chinese politics to developalternative indicators of coercive capacity and to theorize the relationship betweencoercive capacity and Chinarsquos authoritarian resilience more carefully

AcknowledgementThe author wishes to thank Jingkai He Jason Kwon and Hao Wang for excellentresearch assistance and Zach Barter Zack Cooper Joe Fewsmith Iain JohnstonVanya Krieckhaus Adam Liff Peter Lorentzen Andy Mertha Carl MinznerVipin Narang Elizabeth Perry Suzanne Scoggins Caitlin Talmadge YuhuaWang Xie Yue audiences at the Association for Asian Studies CornellUniversity Harvard Universityrsquos Program on Global Society and SecurityPrinceton University and the George Washington Universityrsquos Elliott School forhelpful comments and suggestions

Biographical noteSheenaChestnutGreitens is an assistant professorof political science at theUniversityofMissouri She is also anon-resident senior fellowat theCenter forEastAsianPolicyat the Brookings Institution and an associate in research at the Harvard FairbankCenter Her book Dictators and Their Secret Police Coercive Institutions and StateViolence was published in 2016 by Cambridge University Press

摘摘要要 谈到中国国内安全开支增加的时候 分析人士经常会假设开支的增加

是共产党强制能力提高的证据这片文章的论点是 因为理论基础不够明

晰 所以关于中国国内安全开支的结论经常有缺点在这篇文章 笔者用

中国从 1992 年到 2012 年的国内安全开支的分析来挑战一般常识笔者

认为 除了开支的数量以外 中国怎么利用这样的资源以及需要应对什么

样的威胁也值得注意根据笔者的统计分析 中国国内安全开支 1) 并不

算史无前例 2) 在国家预算占的比例不在扩大 3) 跟别的国家相比不算高而且产生的强制能力也不算大笔者还进一步表明在征收国内安全财政收

入的方面 一部分省市会更困难 而且这些省市平常是有更严重的国内威胁

的地方随着国内安全开支的提高 国内安全机关面对的挑战也加剧了很

多最后 国内安全机关政治地位的改善并不意味着他们比较容易控制中

国社会总之 笔者认为这项分析显示的不是中国国内安全机关的实力 而是国内安全机关的局限

关关键键词词 政法委 政法系统 维护稳定 维稳 公安 强制能力 中国国内安全

开支 群体性事件

ReferencesAlbertus Michael and Victor Menaldo 2012 ldquoCoercive capacity and the prospects for democratiza-

tionrdquo Comparative Politics 44(2) 151ndash169

Rethinking Chinarsquos Coercive Capacity 19

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S0305741017001023Downloaded from httpswwwcambridgeorgcore University of Missouri-Columbia on 05 Jul 2017 at 203959 subject to the Cambridge Core

Andreas Peter and Kelly M Greenhill (eds) 2010 Sex Drugs and Body Counts The Politics ofNumbers in Global Crime and Conflict Ithaca NY Cornell University Press

Bellin Eva 2005 ldquoCoercive institutions and coercive leadersrdquo In Marsha Pripstein Posusney andMichele Penner Angrist (eds) Authoritarianism in the Middle East Regimes and ResistanceBoulder CO Lynne Rienner 21ndash41

Biddle Stephen 2006 Military Power Explaining Victory and Defeat in Modern Battle PrincetonNJ Princeton University Press

Blanchard Ben and John Ruwitch 2013 ldquoChina hikes defense budget to spend more on internalsecurityrdquo Reuters 5 March

Blasko Dennis and John F Corbett Jr 1998 ldquoNo more Tiananmens the Peoplersquos Armed Police andstability in China 1997rdquo China Strategic Review Spring 80ndash103

Brooks Risa and Elizabeth Stanley (eds) 2007 Creating Military Power The Sources of MilitaryEffectiveness Stanford CA Stanford University Press

Brownlee Jason Tarek Masoud and Andrew Reynolds 2015 The Arab Spring Pathways ofRepression and Reform Oxford Oxford University Press

Buckley Chris 2011 ldquoChina internal security jumps past army budgetrdquo Reuters 5 MarchBuckley Chris 2012a ldquoChinarsquos domestic security spending rises to $111 billionrdquo Reuters 5 MarchBuckley Chris 2012b ldquoExclusive China considers downgrading domestic security tsar in next

line-uprdquo Reuters 29 AugustBureau of Justice Statistics 2013 ldquoLocal policerdquo 6 May httpbjsgovindexcfmty=tpamptid=71Chen Xi 2013 ldquoThe rising cost of stabilityrdquo Journal of Democracy 42(1) 57ndash64Cheung Tai Ming 1996 ldquoGuarding Chinarsquos domestic front line the Peoplersquos Armed Police and

Chinarsquos stabilityrdquo The China Quarterly 146 525ndash547Cooper Julian 2014 ldquoThe funding of the power agencies of the Russian state an update 2005 to

2014 and beyondrdquo Journal of Power Institutions in Post-Soviet Societies 16 httppipssrevuesorg4063

Deng Xuan 2011 ldquoCaizheng zhichu guimo jiegou yu chengxiang shouru bupingdeng yigeiziZhongguo shengji mianban shujude shizheng fenxirdquo (Financial expenditure and ruralndashurbaninequality evidence from Chinese inter-province panel data analysis) Jingji pinglun 4 63ndash69

Deng Yanhua and Kevin J OrsquoBrien 2013 ldquoRelational repression in China using social ties todemobilize protestersrdquo The China Quarterly 215 533ndash552

DHS (Department of Homeland Security) 2013 ldquoSecretary Napolitano announces fiscal year 2014budget requestrdquo 10 April httpwwwdhsgovnews20130410secretary-napolitano-announces-fiscal-year-2014-budget-request

DNI (Director of National Intelligence) 2013 ldquoDNI releases budget figure for 2013 NationalIntelligence Programrdquo 30 October httpwwwdnigovindexphpnewsroompress-releases191-press-releases-2013957-dni-releases-budget-figure-for-2013-national-intelligence-program

DOD (Department of Defense) 2013 ldquoDoD releases fiscal year 2014 budget proposalrdquo 10 Aprilhttpwwwdefensegovreleasesreleaseaspxreleaseid=15921

DOJ (Department of Justice) 2013 ldquoDepartment of Justice FY2014 budget requestrdquo 10 April httpwwwjusticegovopapr2013April13-ag-413html

DOJ 2014 ldquoUS Department of Justice FY2014 budget requestrdquo 19 January httpwwwjusticegovjmd2014factsheetsprisons-detentionpdf

Erickson Andrew and Gabe Collins 2013 ldquoNew fleet on the block Chinarsquos coast guard comestogetherrdquo Wall Street Journal China Real Time 11 March httpblogswsjcomchinarealtime20130311new-fleet-on-the-block-chinas-coast-guard-comes-together

Erickson Andrew and Adam Liff 2016 ldquoInstalling a safety on the lsquoloaded gunrsquo Chinarsquos institutionalreforms National Security Commission and Sino-Japanese crisis (in)stabilityrdquo Journal ofContemporary China 25 98 197ndash215

FBI (Federal Bureau of Investigation) 2014 ldquoFederal Bureau of Investigation FY 2014 budgetrequest at a glancerdquo 28 January httpwwwjusticegovjmd2014summarypdffbipdf

20 The China Quarterly pp 1ndash24

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S0305741017001023Downloaded from httpswwwcambridgeorgcore University of Missouri-Columbia on 05 Jul 2017 at 203959 subject to the Cambridge Core

Fewsmith Joseph 2016 ldquoChinarsquos political ecology and the fight against corruptionrdquo ChinaLeadership Monitor 46 httpwwwhooverorgsitesdefaultfilesresearchdocsclm46jfpdf

Fischer Andrew M 2015 ldquoSubsidizing Tibet an interprovincial comparison of western China up tothe end of the HundashWen erardquo The China Quarterly 221 73ndash99

Forsythe Michael 2014 ldquoChina to ramp up military spendingrdquo New York Times 3 FebruaryFravel M Taylor 2007 ldquoSecuring borders Chinarsquos doctrine and force structure for frontier defenserdquo

Journal of Strategic Studies 30(4ndash5) 705ndash737Goldstein Lyle J 2010 Five Dragons Stirring up the Sea Challenge and Opportunity in Chinarsquos

Improving Maritime Enforcement Capabilities Newport RI Naval War College ChinaMaritime Studies Institute

Grauer Ryan and Michael C Horowitz 2012 ldquoWhat determines military victory Testing the mod-ern systemrdquo Security Studies 21(1) 83ndash112

Greitens Sheena Chestnut 2016 Dictators and Their Secret Police Coercive Institutions and StateViolence Cambridge Cambridge University Press

Grieco Joseph M 1993 ldquoAnarchy and the limits of cooperationrdquo In David A Baldwin (ed)Neorealism and Neoliberalism New York Columbia University Press 116ndash140

Guo Gang 2012 ldquoPolitics of local law enforcement spending in Chinardquo Paper presented at theAPSA annual meeting New Orleans 1 September

Guo Xuezhi 2012 Chinarsquos Security State Philosophy Evolution and Politics New York CambridgeUniversity Press

He Qinglian 2012 ldquoChinarsquos stability maintenance system faces financial pressurerdquo China RightsForum December httpwwwhrichinaorgencrfarticle6415

Horowitz Michael 2010 The Diffusion of Military Power Causes and Consequences for InternationalPolitics Princeton NJ Princeton University Press

Hu Wannian 2009 ldquoJingcha xinli jiankang wenti de chengyin jiqi yingdui zhi cerdquo (Causes and policyregarding police mental health problems) Gongrsquoan yanjiu 3 59ndash64

Hudson John 2014 ldquoIn reversal FBI now emphasizes role in law enforcementrdquo Foreign Policy 17January

IISS (International Institute for Strategic Studies) 2001ndash2012 The Military Balance London IISSJakobson Linda 2014 Chinarsquos Unpredictable Maritime Security Actors Sydney Lowy InstituteJiang Steven 2015 ldquoZhou Yongkang from apex of power to lsquocaged tigerrsquo in Chinardquo CNN 11 JuneJohnston Alastair Iain 2012 ldquoWhat (if anything) does East Asia tell us about IR theoryrdquo Annual

Review of Political Science 15 53ndash78JohnstonAlastair Iain 2013 ldquoHownew is Chinarsquos new assertivenessrdquo International Security 37(4) 7ndash48Kan Karita 2013 ldquoWhither weiwen Stability maintenance in the 18th Party Congressrdquo China

Perspectives 1 87ndash93King Gary Jennifer Pan and Margaret Roberts 2014 ldquoReverse engineering censorship in Chinardquo

Science 345(6199) 1ndash7Kyckelhahn Tracy 2012 ldquoState prisons expenditures FY 1982ndash2010rdquo Bureau of Justice Statistics

Department of Justice NCJ239672 December httpwwwbjsgovcontentpubpdfscefy8210pdfLamptonDavidM 1987a ldquoChinese politics the bargaining treadmillrdquo Issues and Studies 23(1) 11ndash41Lampton David M 1987b Policy Implementation in Post-Mao China Berkeley CA University of

California PressLampton David M 2015 ldquoXi Jinping and the National Security Commission policy coordination

and political powerrdquo Journal of Contemporary China 24(95) 759ndash777Lieberthal Kenneth G and David M Lampton (eds) 1992 Bureaucracy Politics and Decision

Making in Post-Mao China Berkeley CA University of California PressLieberthal Kenneth G and Michel C Oksenberg 1988 Policy Making in China Leaders

Structures and Processes Princeton NJ Princeton University PressLiff Adam P and Andrew S Erickson 2013 ldquoChinarsquos defence spending less mysterious in the

aggregaterdquo The China Quarterly 216 805ndash830

Rethinking Chinarsquos Coercive Capacity 21

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S0305741017001023Downloaded from httpswwwcambridgeorgcore University of Missouri-Columbia on 05 Jul 2017 at 203959 subject to the Cambridge Core

Luuml Xiaobo and Pierre Landry 2014 ldquoShow me the money interjurisdictional political competitionand fiscal extraction in Chinardquo American Political Science Review 108(3) 706ndash722

Martinson Ryan D 2014 ldquoThe militarization of Chinarsquos coast guardrdquo The Diplomat 21 NovemberMertha Andrew 2009 ldquoFragmented authoritarianism 20 political pluralization in the Chinese pol-

icy processrdquo The China Quarterly 200 995ndash1012MOF (Ministry of Finance) 1992ndash2002 Zhongguo caizheng nianjian 1992ndash2002 (Finance Yearbook of

China 1992ndash2002) Beijing China State FinanceMOF 1996ndash2009 Difang caizheng tongji ziliao 1996ndash2009 (Local Financial Statistics 1996ndash2009)

Beijing China Financial and Economic PublishingNarang Vipin and Caitlin Talmadge 2017 ldquoCivilndashmilitary pathologies and defeat in war tests using

new datardquo Journal of Conflict Resolution (forthcoming)Nathan Andrew J 2003 ldquoAuthoritarian resiliencerdquo Journal of Democracy 14(1) 6ndash17NBS (National Bureau of Statistics of China) 2003ndash2013 Zhongguo tongji nianjian 2003ndash2013

(Statistical Yearbook of China 2003ndash2013) Beijing China Statistics PressOrsquoBrien Kevin 2011 ldquoStudying Chinese politics in an age of specializationrdquo Journal of

Contemporary China 20(71) 535ndash541Pei Minxin 2012 ldquoIs CCP rule fragile or resilientrdquo Journal of Democracy 23(1) 27ndash41Qian Gang 2012 ldquoPreserving stabilityrdquo China Media Project 14 September httpcmphkuhk

2012091427074Reaves Brian A 2010 ldquoLocal police departments 2007rdquo Department of Justice NCJ 231174 2

December httpwwwbjsgovindexcfmty=pbdetailampiid=1750Reaves Brian A 2011 ldquoCensus of state and local law enforcement agencies 2008rdquo Department of

Justice NCJ 233982 July httpwwwbjsgovcontentpubpdfcsllea08pdfReny Marie-Eve 2011 ldquoWhat happened to the study of China in comparative politicsrdquo Journal of

East Asian Studies 11(1) 105ndash135Roessler Philip 2011 ldquoThe enemy within personal rule coups and civil war in Africardquo World

Politics 63(2) 300ndash346Ruwitch John 2012 ldquoAs Chinarsquos clout grows sea policy proves unfathomablerdquo Reuters 9

DecemberScoggins Suzanne 2016 ldquoPolicing China Struggles of Law Order and Organizationrdquo PhD diss

University of California BerkeleyScoggins Suzanne and Kevin OrsquoBrien 2016 ldquoChinarsquos unhappy policerdquo Asian Survey 56(2) 225ndash242SDRG (Social Development Research Group) 2010 Yi liyi biaoda zhiduhua shixian shehui de chang-

zhijiursquoan (Institutionalize Interest Representation to Realize Long-Term Social Stability) BeijingTsinghua University

Sheen Seongho 2013 ldquoNortheast Asiarsquos aging population and regional security lsquodemographicpeacersquordquo Asian Survey 53(2) 292ndash318

Shi Jiangtao 2012 ldquoBeijing begins big political reshufflerdquo South China Morning Post 20 NovemberShi Xiaochen and Haibo Zhang 2015 ldquoZhongguo difang zhengfu gonggong anchuan zhichu xiaolu

yanjiu yijiyu DEA-Tobit de erjieduan fenxirdquo (Study on the efficiency of public security expenditureby Chinarsquos local government evidence from second-stage DEA-Tobit model) Dianzi keji daxuexuebao (shehui kexueban) 17(1) 12ndash17 22

Shirk Susan 2007 Fragile Superpower New York Oxford University PressSkocpol Theda 1979 States and Social Revolutions Cambridge Cambridge University PressSolomon Peter H Jr 2007 ldquoCourts and judges in authoritarian regimesrdquo World Politics 60(1)

122ndash145State Council Information Office 2012 ldquoMedical and health services in Chinardquo httpwww

china-embassyorgengztbpst1001641htm Accessed 2 June 2017Svolik Milan 2012 The Politics of Authoritarian Rule Cambridge Cambridge University PressTalmadge Caitlin 2015 The Dictatorrsquos Army Battlefield Effectiveness in Authoritarian Regimes

Ithaca NY Cornell University Press

22 The China Quarterly pp 1ndash24

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S0305741017001023Downloaded from httpswwwcambridgeorgcore University of Missouri-Columbia on 05 Jul 2017 at 203959 subject to the Cambridge Core

Tanner Murray Scot 2002 ldquoThe institutional lessons of disaster reorganizing Chinarsquos PeoplersquosArmed Police after Tiananmenrdquo In James Mulvenon (ed) The Peoplersquos Liberation Army asOrganization Washington DC RAND 587ndash635

Tanner Murray Scot and Eric Green 2007 ldquoPrincipals and secret agents central versus local controlover policing and obstacles to lsquorule of lawrsquo in Chinardquo The China Quarterly 191 644ndash670

Wallace Jeremy 2014 ldquoJuking the stats Authoritarian information problems in Chinardquo BritishJournal of Political Science 46(1) 11ndash29

Wang Yuhua 2014a ldquoCoercive capacity and the durability of the Chinese Communist staterdquoCommunist and Post-Communist Studies 47(1) 13ndash25

Wang Yuhua 2014b ldquoEmpowering the police how China manages its coercive leadersrdquo The ChinaQuarterly 219 625ndash648

Wang Yuhua and Carl Minzner 2015 ldquoThe rise of the Chinese security staterdquo The China Quarterly222 339ndash359

Whiting Susan 2004 ldquoThe cadre evaluation system at the grass roots the paradox of Party rulerdquo InBarry Naughton and Dali L Yang (eds) Holding China Together Diversity and NationalIntegration in the Post-Deng Era Cambridge Cambridge University Press 101ndash119

Wines Michael 2009 ldquoChina approves law governing armed police forcerdquo New York Times 27August

Xiao Tiefeng 2013 ldquoMisconceptions about Chinarsquos growth inmilitary spendingrdquoCarnegie Endowmentfor International Peace 28 May httpcarnegieendowmentorg20130528misconceptions-about-china-s-growth-in-military-spendingg76a

Xie Yue 2012 ldquoThe political logic of weiwen in contemporary Chinardquo Issues and Studies 48(3) 1ndash41Xie Yue 2013a ldquoRising central spending on public security and the dilemma facing grassroots offi-

cials in Chinardquo Journal of Current Chinese Affairs 42(2) 79ndash109Xie Yue 2013b Weiwende zhengzhi luoji (Political Logic of Weiwen) Hong Kong Tsinghua

BookstoreXie Yue and Dongsheng Dang 2013 ldquolsquoWeiwenrsquo de jixiao kunjing gonggong anquan kaizhi shijiaordquo

(The efficiency dilemma of ldquoweiwenrdquo from the perspective of public security expenditure) Tongjidaxue xuebao (shehui kexueban) 6 90ndash100

Xie Yue and Wei Shan 2013 ldquoChina struggles to maintain stability strengthening its public securityapparatusrdquo In Wang Gungwu and Zheng Yongnian (eds) China Development and GovernanceSingapore World Scientific 55ndash62

Yao Ziguo 2004 ldquoShilun jianli he wanshan gongrsquoan jingfei baozhang jizhirdquo (Essay on establishingand improving public security budget spending) Gongrsquoan yanjiu 12 78ndash81

Ye Yongguang 2006 ldquoDui gongrsquoan jingfei baozhang biaozhunde ruogan sikaordquo (Thoughts on secur-ing police budget) Gongrsquoan yanjiu 1 66ndash68 75

Yuen Wei Hai Samson 2014 ldquoThe politics of weiwen stability as a source of legitimacy inpost-Tiananmen Chinardquo ECPR Working Paper httpsecpreuFilestorePaperProposal94e6e2c5-c31a-4cd8-8759-266d21ccb81dpdf

Zhong Lena 2009 ldquoCommunity policing in China old wine in new bottlesrdquo Police Practice andResearch 10(2) 157ndash169

Zhu Qing and Xiaohu Wang 2011 ldquoStructural changes of public expenditures in Chinardquo Journal ofPublic Budgeting Accounting and Financial Management 23(4) 569ndash587

Appendix

Calculating US domestic security spending

Comparing domestic security spending in the US and PRC is difficult for severalreasons First Americarsquos federal structure and decentralized policing make

Rethinking Chinarsquos Coercive Capacity 23

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S0305741017001023Downloaded from httpswwwcambridgeorgcore University of Missouri-Columbia on 05 Jul 2017 at 203959 subject to the Cambridge Core

aggregate estimation of US expenditure difficult (Of the sim$155 billion in domes-tic security spending in 2013 around one-third was federal expendituretwo-thirds was state-level prison spending and statelocal police forces)Second an ideal analysis would separate political policing aimed at keeping aparticular regime in power and therefore specific (with gradations) to autocracyfrom the law enforcement tasks that are common to both democracies and autoc-racies Available data however simply do not allow this separationTo reach the estimates of US spending given in this article several federal

departments (including the Department of Homeland Security and parts of theDepartment of Justice such as the Federal Bureau of Investigation and theFederal Bureau of Prisons) were aggregated with state and local spending onpolice courts and prisons Even here judgments on the precise breakdowns ofinternal versus external security are likely to be imperfect the Department ofJustice stated in 2014 that $44 billion of its $276 billion budget went toldquonational securityrdquo including ldquocounterterrorism and counterintelligence pro-grams hellip intelligence gathering and surveillance capabilitiesrdquo while the FBIhas alternately described its mission as either ldquolaw enforcementrdquo or ldquonationalsecurityrdquo or bothCertain categories that were excluded from this aggregation likely make it a

low or conservative one Estimates of state and local police spending in this art-icle do not for example include sheriffsrsquo offices state law enforcement agenciesor special jurisdiction agencies which equal or exceed local police departments innumber45 It is also possible that some portion of the sim$50 billion annual intelli-gence budget should be included in the total but a breakdown of this spending isnot publicly available and therefore is not included46

45 BJS 2013 DOJ 2013 2014 DHS 2013 FBI 2014 Kyckelhahn 2012 Hudson 2014 Reaves 2010 201146 DNI 2013 DOJ 2013

24 The China Quarterly pp 1ndash24

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S0305741017001023Downloaded from httpswwwcambridgeorgcore University of Missouri-Columbia on 05 Jul 2017 at 203959 subject to the Cambridge Core